Struct BakeHyper 

pub struct BakeHyper {

    pub steps_a: usize,
    pub steps_b: usize,
    pub l1_init: f64,
    pub l1_step: f64,
    pub eval_every: usize,
    pub lr_a: f64,
    pub lr_b: f64,
    pub tau: f32,
    pub fcd_layers: usize,
    pub seed: u64,
    pub target_sparsity: f64,
    pub l1_mult: f64,
    pub align: usize,
    pub uniform_inter: bool,
}

Hyper-parameters — defaults are the certified recipe.

Fields

steps_a: usize

steps_b: usize

l1_init: f64

l1_step: f64

eval_every: usize

lr_a: f64

lr_b: f64

tau: f32

fcd_layers: usize

seed: u64

target_sparsity: f64

Target sparsity (0..1). When >0, the best checkpoint must have at least this fraction of neurons pruned; if none qualifies the highest-sparsity checkpoint is used.

l1_mult: f64

L1 aggression multiplier: scales both l1_init and l1_step.

1.0 = harder pruning push, <1.0 = softer.

align: usize

Round each layer’s kept-neuron count UP to a multiple of this (0/1 = off). 32 keeps the defragged FFN on grouped codecs (in % 32 == 0) and SIMD kernels off their scalar tails.

uniform_inter: bool

Force one FFN width across all layers (the max aligned count) — the whole-token GPU graphs require a uniform intermediate size.

Trait Implementations

impl Clone for BakeHyper

fn clone(&self) -> BakeHyper

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for BakeHyper

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for BakeHyper

fn default() -> Self

Returns the “default value” for a type.